Abstracts

Rationale: Magnetic resonance imaging (MRI) has been among the prevailing methods to appraise network compromise in the common epilepsies. In particular, resting-state functional MRI (fMRI) provides insight into the dynamic interactions between distributed regions. When combined with the geometric properties of brain networks, it revealed imbalances of short- and long-range functional connections in temporo-limbic regions ipsilateral to the seizure focus in temporal lobe epilepsy (TLE).¹ However, despite the critical role of the cerebrovascular system in maintaining whole-brain structure and function, it is rarely considered in whole-brain connectivity analyses, thus the link between perfusion-based measurements and network pathology remains a key question in epilepsy. In this study, we mapped cortical perfusion changes in TLE using arterial spin labelling imaging (ASL), and assessed associations to functional connectivity alterations in the condition.
¹Larivière et al., 2020. Epilepsia, 61:1221-1233

Methods: Participants: We obtained high-resolution T1w, resting-state fMRI, and ASL data from 15 adult patients with drug-resistant unilateral TLE (seven males, mean ± SD age = 38.40 ± 10.93 years, 10 left-sided focus, mean ± SD duration = 15.85 ± 11.74 years) and 34 age- and sex-matched healthy controls (16 males, mean ± SD age = 35.79 ± 9.12 years).

Cerebral perfusion: We computed vertex-wise maps of cerebral perfusion and compared TLE patients to controls using surface-based linear models.

Relation to functional network organization: We followed a previous approach¹ and mapped functional connectivity distance across the cortical surface, by combining resting-state fMRI connectivity and geodesic distance analysis. Connectivity distance maps of patients were compared to controls using surface-based linear models. To assess the effects of perfusion on functional connectivity distance, we repeated the analysis while controlling for vertex-wise perfusion measures. Cohen’s d effect sizes were further quantified for both controlled and uncontrolled analyses.
¹Larivière et al., 2020. Epilepsia, 61:1221-1233

Results: Compared to controls, individuals with TLE demonstrated a decrease in cerebral blood flow in temporal lobe regions, inferior parietal lobule regions, precentral gyrus, and postcentral gyrus (p_FWE< 0.1; Figure 1B). Moreover, TLE patients revealed significant connectivity distance reductions in the superior temporal gyrus (p_FWE< 0.05; Figure 2B), which remained robust after correcting for perfusion alterations. Indeed, we observed marginal differences in connectivity reductions patterns prior to and subsequent to the corrections, suggesting independence from underlying vascular abnormalities (p_FWE< 0.05; Figure 2C).